Helm life-cycle hooks/Jobs not run

[qrkourier]

Are you perchance wrapping helm CLI with a something like an Ansible module or TF provider? I've noticed that some wrappers do not fire the hooks unless wait=false, which is not intuitive, and I don't understand why this is necessary. There is a long-running GitHub issue about it.

The workaround is to unset wait and atomic when creating or upgrading the Helm Release. The helm CLI default settings seem to work fine, but when wrapping Helm I've noticed that hooks don't fire without the additional setting. Let me know if you have any ideas for making this more robust.

EDIT: It's starting to make sense why we're seeing what we're seeing. The default for Helm CLI is wait=false, and post-install and post-upgrade don't run when wait=true. It's because those Jobs are only scheduled after the main app pod is "ready," which can't happen until after those Jobs run, so it's a pickle. Maybe those hooks need to be pre-install and pre-upgrade. I don't recall now what forced me to use post-x at the time, but the constraint may have only applied to certain charts, not all. I'll investigate.

re: forum topic - https://openziti.discourse.group/t/deploying-router-on-k8s-via-helm-gives-identity-secret-mount-error/1795/7